On Thursday 29 June 2006 06:59, Justin Erenkrantz wrote:
> What I'd suggest is that we create a branch that has whatever
> API-busting change we want. What's the specific example of something
> that would force a 2.x? Neither errors nor IO abstraction should
> necessarily force 2.x.
I am still secretly hoping that I can get you guys to accept my
apr_proc_mutex_{attach,detach}() patch. I don't know how much that breaks
the API, but I do know that trying to make it not break introduced some
weirdness around the corners.
Dirk
